Best Practice

A teacher’s guide to retrieval practice: Interleaving

Retrieval practice & metacognition
Continuing his series on the potential of retrieval practice, spaced learning, successive relearning, and metacognitive approaches in the classroom, Kristian Still looks at the research underpinning interleaving techniques and draws out some important lessons for teachers


In this series, I am attempting to elaborate and share what the recipe of test-enhanced learning (more commonly known as retrieval practice), spaced learning, interleaving, feedback, metacognition, and motivation might look like in and out of the classroom.

I am reviewing the research and cognitive science behind these concepts and the modulators underpinning the effective retention of knowledge.

In writing this series, nine clear but interlinked elements emerged. I am considering these elements across nine distinct but related articles:

I would urge readers to also listen to a recent episode of the SecEd Podcast (SecEd, 2022) looking at retrieval practice, spaced learning, and interleaving and featuring a practical discussion between myself, teacher Helen Webb, and Dr Tom Perry, who led the Education Endowment Foundation’s Cognitive Science Approaches in the Classroom review (Perry et al, 2021).

This series, in reviewing the evidence-base, seeks to help you reflect on what will work for you, your classroom, and your pupils. This is article four and it focuses on interleaving.


Interleaving

“If your aim is to learn about the character of Sherlock Holmes and the character of Dr Watson then you should block practice. If your aim is to understand the differences in their characters, then interleave.”

I found this explanation from Dr Philip Higham, an expert in cognitive psychology at the University of Southampton, so useful that I made him repeat it three times during our conversation to ensure I got the quote right in my notes.

Also worth dwelling upon is the following from Clare Sealy (2019): “In almost all learning domains, all subjects, all age ranges, substantive knowledge is taught – the building blocks of various subjects. This ‘substance’ is central to being able to think mathematically, or scientifically, or historically, or to communicate clearly.”

An intuitive approach is to focus on learning one topic at a time (e.g. AABB), which cognitive scientists refer to as blocked or massed practice. Its use is consistent with the common assumption that pupils – humans – learn most effectively when topics are introduced in isolation, in sequence – and that repeated practice fosters the development of expertise and fluency after each topic.

It is how many of us were taught and how many of us teach. Indeed, Dedrick et al (2016) report that 78% of practice problems in mathematics textbooks are blocked.

However, researchers have started to investigate an alternative known as interleaved practice or interleaving. Interleaving involves switching between topics (e.g. ABAB), with learners forced to select an appropriate strategy for each topic.

Perry et al (2021) explain: “Interleaving involves sequencing tasks so that learning material is interspersed with slightly (but not completely) different content or activities, as opposed to undertaking tasks through a blocked and consecutive approach. While similar to spaced practice, interleaving involves sequencing tasks or learning content that share some likeness, whereas a spaced practice approach uses intervals that are filled with unrelated activities.”

Kornell and Bjork (2008) famously presented learners with impressionist landscape artwork in blocked or interleaved formats and then tested their ability to identify the artist who had painted novel example paintings.

There were two key findings: first, interleaving was superior to blocking in this test; second, students’ metacognitive awareness of their learning was faulty in that they tended to believe (incorrectly) that they had learned better via blocking (a phenomenon already touched upon in this series).

Both these findings have since been replicated multiple times and generally the research shows that interleaved practice is superior to blocked practice.

More recently, researchers have been more assertive: “We conducted a large randomised classroom study and found that a greater emphasis on interleaved practice dramatically improved test scores” (Rohrer et al, 2019) – and “at no (or very little) additional time cost” (Wiseheart et al, 2019).

Furthermore: “It can boost both memory and transfer and applies across different subject domains.” (Firth et al, 2021).


So, why is blocked practice so popular?

One reason that we persist with blocked practice could be that pupils and teachers (and textbook authors) might not be aware of interleaved practice. Or it might be that they are swayed by the familiarity and fluency of blocked practice (Bjork et al, 2013; Kirk-Johnson et al, 2019), which encourages pupils and teachers to falsely believe that blocked practice enhances efficacy.

Finally, and less provocatively, it might simply be that blocked practice is more convenient, following each lesson with a group of problems devoted to that lesson. Compared to this, “interleaving instruction can feel disorganised and chaotic, and may require a significant time investment to restructure curriculum” (Sana & Yan, 2021).

Let’s not forget, if your teaching materials are not categorised or tagged, then re-ordering learning will be more time-consuming (and rather daunting).


What is the theory behind interleaving?

Several theoretical mechanisms may account for the observed benefits of interleaving. However, these are not necessarily mutually exclusive. It is part of the reason that this area of cognitive psychology is hard to pin down empirically.

Rohrer et al (2019) state: “In addition to any benefits of mixture per se, the interleaving of practice problems in a course or text inherently incorporates the learning strategies of spacing and retrieval practice, each of which is an effective and robust learning strategy.”

Both spacing and repeated retrieval practice enhance the durability and accessibility of knowledge as we discussed in articles two and three. The most prevalent explanations reference two theories as to why this is:

  • Inductive learning, where learners discover rules or concepts by observation and examples (as opposed to deductive learning, where students are given rules that they then need to apply).
  • The discriminative-contrast hypothesis – suggesting pupils learn to discriminate and to contrast problems when those problems are presented in interleaved form.

With blocking, once you know what solution to use the hard part is over – the predictability of blocking obviates any need to engage in strategy selection.

With interleaving, each practice attempt is unique, therefore rote responses are unreliable. Pupils are continuously searching for different solutions and that process improves their ability to learn critical features, processes and concepts, which then better enables pupils to select and execute the correct response.

A second explanation is that interleaving strengthens memory associations. With blocking, a single strategy, temporarily held in short-term memory, is sufficient. That is not the case with interleaving, where the correct solution is constantly changing.

As a result, greater attention is needed, re-encoding contextual cues, with pupils continually required to “retrieve” different responses, repeatedly bringing them into short-term memory. Repeating that process “fosters more relational processing” (Samani & Pan, 2021), reinforcing neural connections between different tasks and correct responses, which enhances learning.

In both cases, the similarity of materials is an obvious modulator. As Chen et al (2021) point out: “Differences between interleaved and blocked practice can be more difficult to obtain depending on the discriminability of the interleaved materials.”

Which basically means that when interleaved topics are “very obviously different” and “immediately discriminable”, “the interleaving effect is unlikely to be obtained” (Chen et al, 2021).

Accordingly, teachers would be advised to present materials that are more difficult to discriminate between – Perry et al’s (2021) “slightly but not completely different content or activities” – in order to leverage an interleaving benefit.

Spaced and interleaved practices have both been identified as effective but they are sometimes conflated as a single strategy, which has led to the erroneous assumption that both effects have the same cause.

To be clear, spaced learning, with rest-from-deliberate-practice, generally results in more improved outcomes than massed practice without rest-from-deliberate-practice.

Interleaved practice also consists of spaced learning, but topics are interleaved rather than involving rest-from-deliberate-practice.

Finally, it is not merely the principle of interleaving that you need to take into consideration but also your learning aims – and so we come back to Sherlock Holmes and Dr Watson.


Interleaving – ‘a bad-tasting cough syrup’

Without doubt, the above is my favourite interleaving reference (from Rohrer, 2012).

In a similar vein, Samani and Pan (2021) concluded: “Despite benefiting more from interleaved practice, students tended to rate the technique as more difficult and incorrectly believed that they learned less from it.

“Thus, in a domain that entails considerable amounts of problem-solving, replacing conventionally arranged with interleaved homework can (despite perceptions to the contrary) foster longer lasting and more generalisable learning.”

In their research exploring the extended effects of interleaving on factual knowledge and problem-solving ability, Samani and Pan replaced the common blocked homework approach with interleaved homework, concluding with a surprise in-class test.

With respect to overall performance, students correctly solved more blocked than interleaved homework problems, confirming their metacognitive judgements.

However, “belying the patterns observed on the homework assignments”, students who had completed interleaved homework “well outperformed” those who had completed blocked homework on the surprise criterial test.

Interleaving improved the ability to correctly “recall and use prior knowledge” in an attempt to generate solutions to novel problems. Importantly, as the weight of content increased (as students progressed through the course), the effect size of the interleaving advantage became larger still.

Samani and Pan were also able to show the effects of interleaving on two distinct sub-measures of test performance: (a) whether students were able to correctly recall necessary formulas (recall memory) and (b) whether students were able to devise and execute a multi-step problem-solving strategy (problem-solving accuracy).

Both spacing and interleaving have been shown to lead to slower and more error-prone learning in the short-term, but to boost memory retention when implemented over a longer time period.

So this, of course, is one of the reasons that pupils, in spite of being told/shown their test scores, still report that they prefer massed practice and re-reading/restudy-style revision strategies.

As Rohrer (2012) observed: “In this scenario, interleaving is like bad-tasting cough syrup – ineffective because children refuse to use it.”

These are issues we will revisit when we discuss illusions of knowing/competence in article eight.


What about interleaved retrieval practice?

In a pre-publication paper, Sana and Yan (2021) posed an interesting question: Can interleaved retrieval practice enhance learning in classrooms?

Across a four-week period, 155 pupils took a 10 to 12-minute weekly quiz in their science courses testing half of the concepts taught that week. Questions on each quiz were either blocked by concept or interleaved with different concepts. A month after the final quiz, pupils were tested on the concepts covered. The pupils performed better on concepts that had been block quizzed (54%) than not quizzed (47%), as we would expect. However, the interleaved quizzes led to even greater scores (63%).


To interleave or not?

When it comes to revision lessons, which is essentially relearning, this is most definitely the time to get your interleaved quizzing going.

Professor Veronica Yan, a cognitive scientist with a professional interest in interleaving, told me: “Exam season is the perfect example of how interleaving and retrieval practice can easily go hand-in-hand. Exams cover a wide range of material and the goal is that students are able to readily recall and apply information that they have learned throughout the year (or indeed the last few years).

“Yes, it requires a curriculum rethink, but we shouldn't be practising this type of retrieval solely during the revision period, rather we should be practising and maintaining knowledge throughout the year.

“This leverages the benefits of spacing (for long-term memory retention), but there are additional benefits that come from the specific juxtaposition of related topics: not least that interleaving gives opportunities for students to notice connections between topics.

“So they are not just strengthening recall, but they are also strengthening their knowledge networks.”

Furthermore, Yan and Sana (2021) raised the valid point that students rarely study only one course at a time. So how should the study of unrelated courses be sequenced? Blocked by course to avoid “unproductive juxtapositions” or interleaved across different courses because it inherently involves spaced practice?

In conversation with Prof Yan, I understood that her conclusions may prove useful for students studying multiple subjects. The optimal schedule, she told me, involves interleaving “at either the concept or the domain level, but not both, nor neither”.

Interleaving is clearly effective for inductive learning of material with high similarity between categories, but not within categories.

So we might interleave Romeo and Juliet with Animal Farm, but we should avoid interleaving two similar Shakespeare plays, for example. Or in your geography lessons, don’t interleave coastal and river landscapes, nor tropical rainforests and boreal forests (too similar), but you could interleave tectonic hazards and urbanising, or tropical rain forests and energy consummation.


Takeaways

  • “The rationale for interleaving is that it may support learners to discriminate between two similar concepts or methods. Teachers may be able to increase the level of challenge presented through interleaved tasks by introducing more similar items of learning.” Perry et al, 2021
  • Present materials that are more difficult to discriminate between if your aim is to leverage an interleaving benefit.
  • When students must learn to distinguish between concepts, interleave rather than block.
  • Break-up longer assessments into more frequent, shorter quizzes.
  • Consider the sequence in which questions are assigned, relearned and revisited.
  • Homeworks and lesson starters are often targeted for short, retrieval practice. Consider interleaving the retrieval items.
  • And if you only have time to read one paper on this topic: Interleaved practice enhances memory and problem-solving ability in undergraduate physics (Samani & Pan, 2021): https://go.nature.com/3GcZSW9


  • Kristian Still is deputy head academic at Boundary Oak School in Fareham. A school leader by day, together with his co-creator Alex Warren, a full-time senior software developer, he is also working with Leeds University and Dr Richard Allen on RememberMore, a project offering resources to teachers and pupils to support personalised spaced retrieval practice. Read his previous articles for SecEd via https://bit.ly/seced-kristianstill


References: For all research references relating to this article, go to https://bit.ly/3vLKQDk

Acknowledgement: This article would not have been possible without the author’s on-going conversations with Helen Pipe, a head of geography and trust geography lead. Helen applied her geographical expertise and interest in interleaving to build a comprehensive RememberMore resource, enabling teachers and pupils to explore 26 Geography Case Studies. Her professional generosity and enthusiasm have been quite humbling.

ResearchED: Kristian will be speaking at the first ever ResearchED Berkshire taking place at Desborough College in Maidenhead on May 7. Visit https://researched.org.uk/event/researched-berkshire/

RememberMore: RememberMore delivers a free, personalised, and adaptive, spaced retrieval practice with feedback. For details, visit www.remembermore.app or try the app and resources via https://classroom.remembermore.app/